Re: oncheck -cI

Final update on this thread (I hope and pray):

Over the weekend I tried to dbexport our database and the dbexport
failed with the following errors when it got to systabauth

*** load table ***

-211 - Cannot read system catalog (systabauth).
-103 - ISAM error: illegal key descriptor (too many parts or too
long).

So IBM's option 1 to fix it went out the window, and we started to
see that the corruption was getting worse as in more rows affected.

This morning when our users started work some of them where getting
access and other strange errors, after a little investigation it seems
the systabauth index corruption had gotten to the point where it was
preventing people from access our tables. So got everyone off the
system and got IBM involved.

IBM Down Systems Department was able to log on and fix the corrupt
indexes. Took IBM over an hour to fix the indexes and a couple of
different attempts but they saved the day. Yaaaaaah!!!!

Things learned for this problem:
1 - When a sys table is involved get IBM support helping out sooner
then later.
2 - Don't try to drop/rebuild any tables matching the tabids that
are
corrupt systabauth index. (will leave records of
the old tabid and
also add the new tabid) - get the indexes fix
first.

John

.

Relevant Pages

  • Re: JFS2 file corruption
    ... namely the final fix for this won't be ... >>fragmented enough the inode table can't grow. ... >>think long and hard before you implement anything jfs2. ... IBM have already checked out ...
    (comp.unix.aix)
  • Re: JES2/SDSF APAR OA24615 & Fixtest PK69442
    ... IBM found that we had a lot of sysout on the output queue with bits ... We ended up putting on the fix for PK52910, modifying the config of an ISV ... when z/OS V1R9 was implemented there was a SDSF ... the sender, and then destroy all copies of this email. ...
    (bit.listserv.ibm-main)
  • Re: Word freezes, then crashes
    ... If this is a document corruption, then saving as HTM does usually remove the corruption. ... But it isn't a 100% fix. ... As the Webrider suggests, the ultimate is to make it all plain text and then insert the text file into a new blank document. ... I have never had a problem since I unchecked those boxes, ...
    (microsoft.public.word.application.errors)
  • Re: JFS2 fundamentally broken
    ... and has been given the opportunity to fix the problem. ... >> Now I mentioned above that IBM had been given the opportunity to fix ... >> situations where readdir() could return duplicates. ... >> the only people complaining, this would not be a priority. ...
    (comp.unix.aix)
  • Re: How many people are using 2.6.16?
    ... as part of the dirty bit cleanups and sanity checkingwe did actually seem to fix a long-standing CIFS corruption. ... We did have one bug report of data corruption in cifs on older kernels copying large files which this resolves, ...
    (Linux-Kernel)